Position Decay attribution model

An attribution model is a method used to determine the impact each marketing channel has on converting visitors into new signups. 

Basically, visitors may find your website from a variety of sources such as social media posts, ads, newsletters, google searches etc. These interactions are called touchpoints. 

Here are the standard attribution models as defined by Google Analytics.

InnerTrends uses an attribution model that is very well suited to SaaS businesses. It is called the Position Decay attribution model and it’s a combination between the Time Decay and the Position based models defined by Google, as it allocates a decreasing number of points to each traffic source according to their position with respect to the moment of conversion.

The Position Decay Attribution Model

The Position Decay is a multi-touch attribution model that uses an algorithm which allocates points to all the traffic sources one used to convert, giving them credit in reverse order. This model assumes that a visitor’s interest in a product grows with every interaction, and the last one is the most important, as it determines the visitor to convert. That’s why the highest score is attributed to the last traffic source; then the score decreases progressively for  every traffic source until the first one. 

How to Calculate the Score Attributed to Each Traffic Source

The maximum number of traffic sources that InnerTrends saves and analyses is the last 7 sources per visitor. And this is the algorithm used to calculate the score attributed to each traffic source:

The sum of all positions = 1 + 2 + 3 + 4 + 5 + 6 + 7 =  28

The attributed score for the:

1st traffic source = 1/28 * 100%  = 4%

2nd traffic source = 2/28 * 100% = 7%

3rd traffic source = 3/28 * 100% = 11%

4th traffic source = 4/28 * 100% = 14%

5th traffic source = 5/28 * 100% = 18%

6th traffic source = 6/28 * 100% = 21%

7th traffic source = 7/28 * 100% = 25%

       ______

        100%

Example

Let’s assume someone is looking for an accounting app. He first uses a search engine to compare a variety of accounting apps and finds your site (this is called organic search). He is then targeted with an AdWords campaign and clicks on one of your ads. After a while he comes back to your site after noticing one of your posts on social media. Your newsletter might finally convince him to sign up.

The Position Decay model, which is used by InnerTrends, will give most of the credit to the newsletter, with a percentage of 32%. Your social media efforts will be allocated 27%. Your AdWords campaign will receive 23%, while the organic search will get the smallest score 18%.

Because there are only 4 traffic sources in this particular case, this is how the percentages have been calculated:

The sum of all positions = 4 + 5 + 6 + 7 =  22

The attributed score for the:

1st traffic source = 4/22 * 100% = 18%

2nd traffic source = 5/22 * 100% = 23%

3rd traffic source = 6/22 * 100% = 27%

4th traffic source = 7/22 * 100% = 32%

                   _____

                   100%